AI in Education
The use of generative AI (GAI) in education has been a hot topic as of late. Large universities such as the University of Sydney, the University of New South Wales and Macquarie University have made drastic moves to embrace GAI as a part of their teaching and research, and Education NSW has rolled out a custom-made GAI chatbot to be used in public schools. However, there are serious concerns about its value and impact on the quality of education.
The main concern about the use of GAI in education is ‘cognitive offloading’: the process of giving mental tasks to an external entity to complete. The reason cognitive offloading is problematic in learners is that practice is vital to the learning process - if a student never writes an essay without the help of GAI, they will never develop the skills to do so on their own. Another problem with cognitive offloading in learners is the 'safety gap', or the difference in ability between the learner and the tool, which results in them being unable to adequately check the quality of the work. Even in researchers, who one might no longer regard as ‘learners’, experience decreased research quality and resilience from task fatigue. Meanwhile, teachers report what has been described as ‘AI addiction’, but may better be described as AI dependency, as a result of integrating it too heavily into their workflow. Since learning and maintaining abilities requires struggle and friction, the ease which GAI provides is deleterious towards the goals of education.
GAI also disproportionately harms vulnerable populations, including Indigenous students and students with disability. Broadly, GAI has a tendency towards perpetuating predominant narratives, which includes disparaging and misinformed narratives about minority populations. More specifically, for Indigenous peoples, GAI goes against many of the core principles of their pedagogy, including situatedness, relatedness and kinship. For students with disability, evidence demonstrates that GAI does a poor job of adequately accommodating and responding to diverse needs. In particular, the text-based conversational format is inaccessible to many, leading to exclusion and decreased autonomy.
Any educational value of GAI must ultimately be weighed against its other detrimental impacts, including on the environment, mental health, workers’ rights and Indigenous communities. If GAI was a revolutionary, high-impact tool, the argument could be made that the benefits outweigh the risks. However, given that no benefits have been sufficiently supported by evidence, a large-scale or general-purpose rollout appears to be unjustified. To put it in more personal, direct terms: we as educators do not have the right to inflict harm onto others, directly or otherwise, to make our jobs easier.
